AVHRR notes

What is the thermal infrared emissivity of ice?

About 0.98.

• Warren (1982) reports that the "infrared emissivity of snow is quite insensitive to snowpack parameters. For many purposes one can simply assume a snow emissivity of about 99%, as was measured by Griggs [1968]".
• Hobbs (1974, p. 223) devotes part of one page to the subject. "Since the absorptivity of ice is large for wave numbers below 10 000 cm-1, and the reflectivity is small, the emissivity of ice should be close to unity in the infrared region"
• Rao, et al. (1990, p. 30-32) provide a few, primarily qaulitative comments on snow and ice emissivity. "...emisivity appears to decrease as the frequency increases, unlike most other natural surfaces. As sea ice ages, bubbles, racks and other metamophosing processes give an unsmooth texture and create scattering centers. Emissivities of heterogeneous substances (multiyear sea ice and dry snow) which contain scatterers have been computed by Poe et al. (1972), Hoffer and Matzler, (1980), and Chang et al. (1976). Since the depth of MW penetration increases with decreasing frequency, multspectral measurements have been used for determining the subsurface characteristics of sea ice (Hollinger et al., 1984; Cavalieri et al., 1983). The analysis becomes more complex when stratification of different surfaces occurs, e.g., snow over ice, or ice over water (England, 1975). Identification of stratified ice and snow layers also may be possible from polarization measurements."
  
  Rao et al. present a graph which shows 'lake and new sea ice' to have emissivities of approximately 0.85 between 10 and 37 GHz, increasing slightly with frequency. (No citation?)